Slim and robust eye tracker on eyeglass temples with NIR patterned mirrors

Area of Science:

• Human-Computer Interaction
• Wearable Technology
• Computer Vision

Background:

• Eye trackers are vital for advanced display systems like head-mounted displays (HMDs) and augmented reality (AR) glasses.
• Existing eye trackers face challenges in accuracy and robustness due to space constraints for components in compact devices.
• Developing unobtrusive and accurate gaze estimation solutions is critical for seamless AR/VR integration.

Purpose of the Study:

• To propose a novel, slim eye tracker design integrated into foldable temples for HMDs and AR glasses.
• To enable accurate gaze estimation while maintaining a compact form-factor and unobstructed user vision.
• To present a new calibration and gaze extraction method optimized for the temple-mounted system.

Main Methods:

• Developed a temple-mounted eye tracker utilizing a near-infrared (NIR) imaging system.
• Incorporated a patterned NIR mirror for calibration markers.
• Introduced a unique calibration and gaze extraction algorithm considering the mirror's spatial reflectance properties.
• Evaluated gaze extraction accuracy with multiple users in realistic scenarios.

Main Results:

• Demonstrated accurate gaze estimation capabilities with the novel temple-mounted eye tracker design.
• Wearable prototypes were successfully developed and tested.
• The proposed algorithm effectively utilized the patterned mirror for calibration and gaze extraction.
• User studies confirmed the system's performance in realistic usage conditions.

Conclusions:

• The temple-mounted eye tracker offers a viable solution for accurate gaze estimation in space-constrained wearable devices.
• The design facilitates slim form-factors and unobstructed vision, enhancing user experience in AR/VR.
• The developed calibration and gaze extraction algorithm shows significant applicability for future eye-tracking systems.